Predicting the Dead Oil Viscosity of Reservoir Fluids: a case study of the Niger Delta

Abstract

Knowledge of the Dead-oil viscosity is one of the most important factors in developing viscosity empirical correlations because other viscosities are obtained from it. A better accuracy of its model will enhance the accuracy of the overall viscosity parameters. This paper developed a new dead oil viscosity using laboratory PVT experimental data obtained from different oil wells across the Niger Delta. The Model was developed starting with the knowledge of Linear and Non-Linear Multiple regression analyses. A Matrix system was adopted to solve the experimental equations through MATLAB in conjunction with Microsoft Excel Application. Multiple statistical analyses were adopted to ascertain the accuracy of the derived model. The empirical correlation gave a Percent Mean Relative Error of -6.67776, Percent Mean Absolute Relative Error of 25.55437 and a 0.6728 Coefficient of Correlation. It also gave better performance plot, compared with those of existing correlations.Keywords: Dead oil viscosity, Pressure, Volume, Temperature, Multiple statistical analyses.